A model for multivariate longitudinal rank data with application to glioma patients

This paper proposes a model to analyze longitudinal rank responses using a Bayesian approach with a random effects framework‎. ‎We consider rank responses that are implicitly determined by their latent variables‎. ‎Further‎, ‎the usual univariate model‎, ‎as well as a multivariate model‎, ‎is also considered for analyzing the multiple longitudinal rank responses‎. ‎We use random effect vectors to evaluate the correlation between individual responses across time‎. ‎Also‎, ‎a Bayesian approach that is used to yield Bayesian estimates of the model's parameters‎. ‎Some simulation studies are conducted to estimate the parameters of the considered models‎. ‎The model is used for a neurocognitive data set of Glioma patients who underwent surgery‎. ‎The results of the data analysis are presented to illustrate the method.

کلیدواژه ها:

Latent variable ، Longitudinal rank data ، Neurocognitive data set ، Random effect